A model of a time-sharing system with two classes of processes

  • Alexandre Brandwajn
Part of the Lecture Notes in Computer Science book series (LNCS, volume 34)


We present a model of a multiprogrammed, virtual memory interactive system, in which the processes are assumed to form two different classes as regards characteristics such as total compute time, input-output rate and program locality. The effect of memory sharing among processes is explicitly taken into account via life-time functions. We use our model to examine the efficiency of two policies of controlling the admission of processes into real core in order to avoid thrashing with two fixed-partition memory allocation schemes. An approximate analytical solution for our model is obtained owing to an equivalence and decomposition approach.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [1]
    Scherr, A.L.: An analysis of time-shared computer systems. Cambridge (Mass.): M.I.T. Press 1967.Google Scholar
  2. [2]
    Arora, S.R., Gallo, A.: The optimal organization of multiprogrammed multilevel memory. Proc. ACM-SIGOPS Workshop on System Performance Evaluation, Harvard University, April 1971, p. 104–141.Google Scholar
  3. [3]
    Buzen, J.: Queueing network models of multiprogramming. Harvard University, Ph. D. Thesis, 1971.Google Scholar
  4. [4]
    Moore, C.G.: Network models for large scale time-sharing systems. University of Michigan, Ph. D. Thesis, 1971.Google Scholar
  5. [5]
    Sekino, A.: Throughput analysis of multiprogrammed virtual-memory computer systems. 1st Annual SICME Symposium on Measurement and Evaluation, Palo Alto, 1973, p. 47–53.Google Scholar
  6. [6]
    Brandwajn, A., Buzen, J., Gelenbe, E., Potier, D.: A model of performance for virtual memory systems. Proc. 1974 SIGMETRICS Symposium, October 1974.Google Scholar
  7. [7]
    Jackson, J.R.: Jobshop-like queueing systems. Management Science 10, 131–142 (1963).Google Scholar
  8. [8]
    Buzen, J.P.: Computational algorithms for closed queueing networks with exponential servers. Comm. ACM 16, 527–531 (1973).Google Scholar
  9. [9]
    Muntz, R.R., Baskett, F.: Open, closed and mixed networks of queues with different classes of customers. Stanford Electronic Laboratories, Technical Report No. 33, August 1972.Google Scholar
  10. [10]
    Courtois, P.: On the near-complete-decomposability of networks of queues and of stochastic models of multiprogramming computing systems. Carnegie Melon University, November, 1971.Google Scholar
  11. [11]
    Avi-Itzhak, B., Heyman, D.F.: Approximate queueing models for multiprogramming computer systems. Operations Research 21, 1212–1230 (1973).Google Scholar
  12. [12]
    Mitrani, I.: Nonpriority multiprogramming systems under heavy demand conditions — customers' viewpoint. Journal ACM 19, 445–452 (1972).Google Scholar
  13. [13]
    Chandy, K.M., Herzog, U., Woo, L.: Approximate analysis of general queueing networks. IBM Research Report RC 4931, July 1974.Google Scholar
  14. [14]
    Brandwajn, A.: A model of a time-sharing virtual memory system solved using equivalence and decomposition methods. Acta Informatica 4, 11–47 (1974).Google Scholar
  15. [15]
    Sauer, C.H., Chandy, K.M.: Approximate analysis of central server models with non-exponential service distributions, different classes of customers and priority queueing disciplines. University of Texas at Austin Report TR-45 (1974).Google Scholar
  16. [16]
    Denning, P.J., Graham, G.S.: Multiprogrammed memory management. To appear in IEEE Proceeding on Interactive Computer Systems, June 1975.Google Scholar
  17. [17]
    Belady, L., Kuehner, C.J.: Dynamic space sharing in computer systems. Comm. ACM 12, 282–288 (1969).Google Scholar
  18. [18]
    Chamberlin, D., Fuller, S., Liu, L.: An analysis of page allocation stategies for multiprogramming systems with virtual memory. IBM J. R and D, September 1973.Google Scholar
  19. [19]
    Simon, H. Ando, A.: Aggregation of variables in dynamic systems. Econometrica 20, 111–138 (1961).Google Scholar
  20. [20]
    Ando, A., Fisher, F.M.: Near-decomposability, partition and aggregation, and the relevance of stability discussions. International Economic Review 4, 53–67 (1963).Google Scholar
  21. [21]
    Little, J.D.: A proof of the queueing formula L = λW. Operations Research 9, 383–387 (1961).Google Scholar
  22. [22]
    Jordan, C.: Calculus of finite differences. New York: Chelsea Publishing Company 1965.Google Scholar
  23. [23]
    Brandwajn, A.: Equivalence et décomposition dans les modèles à files d'attente et leur application à l'évaluation des performances de systèmes d'exploitation. Thèse d'Etat, University of Paris VI, January 1975.Google Scholar
  24. [24]
    Brandwajn, A.: A model of an interactive system with two classes of processes. To appear.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1975

Authors and Affiliations

  • Alexandre Brandwajn
    • 1
  1. 1.IRIA-LABORIA Domaine de Voluceau RocquencourtLe ChesnayFrance

Personalised recommendations